The Split.io SDK API Python client works with Flask out of the box. Both our in-memory and Redis backed clients work well with Flask.

The following section shows how to use Split.io in a simple one-view Flask app.

This example assumes that the Split.io configuration is save in a file called splitio-config.json.

import logging
from flask import Flask, render_template, request

from splitio import get_factory

logging.basicConfig(level=logging.INFO)

app = Flask(__name__)

factory = get_factory('SOME-API-KEY', config_file='splitio-config.json')
# Redis: The redisHost key must be set into config-file
# In-memory: Don't set redis configurations
client = factory.client()


@app.route('/')
def index():
    user = request.args.get('user', '')

    context['some_treatment'] = client.get_treatment(user, 'some_feature')
    context['some_other_treatment'] = client.get_treatment(user, 'some_other_feature',
        {'number_attribute': 42, 'date_attribute': 1466185587010})

    return render_template('index.html', **context)

When using the Redis client the update scripts need to be run periodically, otherwise there won't be any data available to the client.

As mentioned before, if the API key is set to 'localhost' a localhost environment client is generated and no connections to Split.io are made as everything is read from .split file (you can read about this feature in the Localhost Environment section of the :doc:`/introduction`.)